There is a fundamental gap in our knowledge of how antibiotic resistant bacteria emerge and spread, and the ecological factors that favor their success. The continued existence of this gap is an important problem in devising better strategies of antibiotic stewardship and targeting them at the population - either host or bacteria - where they will have the greatest effect. The long-term goal of this application is the development of such strategies. Tracking resistance as it emerges, including ultimately unsuccessful variants that lead to evolutionary dead-ends, is a powerful way of identifying lineages and features associated with success. The objective of this application is to study the re-emergence of resistance in the pneumococcus following implementation of an effective vaccine that targets the great majority of resistant strains of this pathogen. The central hypothesis is that emerging resistant strains will have been present, though rare, before vaccination, and as they become more common they will come to predominate in regions with a large amount of antibiotic use. The applicants have formulated this hypothesis on the basis of preliminary data collected in their laboratories, and examination of data following previous vaccine introductions. The rationale for the proposal is that vaccination offers an opportunity to observe the emergence of resistance, and relate it both to features of the environments in which it emerges and the properties of the emerging strains themselves. The hypothesis will be addressed with three specific aims: 1) Define the origins of emerging resistant clones and their relationship to pre-existing populations; 2) Define ecological factors associated with the emergence of resistance in different regions; and 3) Define genomic properties of successful clones. Under the first aim, bacteria will be characterized by whole genome sequencing, with methods already in use in the PI's laboratory, will be used to determine the relationship between the prevaccine pneumococcal population, and resistant strains sampled by collaborators from a population of more than 29 million persons. The second aim will ask whether differences between regions in the prevalence of resistance, and the exact strains present, are best explained by differences in antibiotic use or other factors, applying methods developed by the co-I. The third aim will examine the genomes gathered in aim 1, to test for properties that could explain the success or failure of strains. The expectation is that these will reflect a history of recombination, which shuffles existing genetic material into novel combinations and which the PI has previously implicated in the spread of resistance in this pathogen. This approach is innovative in its combination of ecological and genomic analyses, and its emphasis on observing resistance as it emerges. The research is significant because it is expected to vertically advance understanding of the epidemiology of resistance in pneumococcus and other pathogens, and suggest means by which it may be limited through careful antibiotic stewardship.